I live at the intersection of local optimization and dynamic experiences, in everything from graphics to LLM inference.
- Summit, NJ
-
19:35
(UTC -04:00)
Highlights
- Pro
Pinned Loading
-
spoke
spoke PublicVoice-first macOS control surface for dictation, assistant workflows, and hands-free control.
Python
-
mlx-quant-toolkit
mlx-quant-toolkit PublicLocal-first Python pipeline for analyzing Mixture-of-Experts (MoE) expert weight matrices from .safetensors and .npz, computing per-expert stats, and optionally simulating MLX quantization/dequanti…
Python 1
-
inferoscope
inferoscope Publicinferoscope is a Python toolkit for capturing completed-token MoE routing events, validating them into durable run bundles, and preserving a clean boundary between raw model behavior and any later …
Python
-
Something went wrong, please refresh the page to try again.
If the problem persists, check the GitHub status page or contact support.
If the problem persists, check the GitHub status page or contact support.




